InsightFinder closed a $15 million Series B round led by Yu Galaxy on April 16, 2026, bringing total funding to $35 million. The company is building an AI agent that watches over other AI agents, and the timing maps precisely to a market that just received its first empirical data on how badly enterprise AI agent deployments are failing.
The company’s product, Autonomous Reliability Insights (ARI), uses unsupervised machine learning, proprietary large and small language models, predictive AI, and causal inference to monitor entire enterprise technology stacks. Not just AI models. Infrastructure, data pipelines, compute, networking, and the AI systems running on top of all of it, treated as a single interconnected system, according to TechCrunch.
The Full-Stack Diagnosis Problem
CEO Helen Gu, a computer science professor at North Carolina State University who previously worked at IBM and Google, framed the core problem to TechCrunch: “In order to diagnose these AI model problems, you need to actually monitor and analyze the data, the model, and the infrastructure together. It’s not always a model problem or a data problem; it’s a combination. Sometimes, it’s simply your infrastructure.”
One example from InsightFinder’s customer base makes this concrete. A major U.S. credit card company’s fraud-detection model was drifting, producing less accurate results over time. Because InsightFinder was monitoring the entire infrastructure stack rather than just the AI model, it identified the root cause: outdated cache in specific server nodes. The model itself was fine. The infrastructure feeding it stale data was the problem, TechCrunch reports.
Without full-stack observability, that failure would have been misdiagnosed as a model quality issue. Engineering teams would have wasted cycles on retraining and fine-tuning while the actual root cause went unfixed.
The Customer Base and Competitive Position
InsightFinder’s current customer roster includes UBS, NBCUniversal, Lenovo, Dell, Google Cloud, and Comcast, according to TechCrunch. The company has been using machine learning for IT infrastructure monitoring since 2016, predating the current AI agent wave by nearly a decade. Gu holds foundational patents in ML and distributed systems monitoring, per the EINPresswire release.
The observability space is crowded. InsightFinder competes against Grafana Labs, Fiddler, Datadog, Dynatrace, New Relic, and BigPanda, all of which are building AI-specific monitoring capabilities. Gu told TechCrunch that retention has been strong: “We actually rarely lose customers to anybody so far. This is about the insights. The problem is that a lot of data scientists understand AI, but they don’t understand the system. And a lot of SRE developers understand the system, but not the AI.”
The Market Context
The Cloud Security Alliance published data on the same day showing that 47% of enterprises have experienced an AI agent security incident in the past year and that detection and response times stretch to hours and days. InsightFinder’s ARI is designed to compress that detection window from hours to minutes by correlating signals across the full stack automatically.
The $35M in total funding positions InsightFinder as a mid-stage entrant in a category that is transitioning from traditional APM (application performance monitoring) to AI-native observability. The question for the category is whether existing APM vendors like Datadog and Dynatrace can retrofit their platforms for AI agent workloads fast enough, or whether specialized startups like InsightFinder with decade-long ML monitoring expertise will own the production AI reliability layer.